In order to meet the needs of intelligent operation and maintenance of distribution networks, a large number of wireless sensors are deployed inside and outside the power grid for digital sensing and immediate control of the grid. However, once these wireless sensors fail, they will generate erroneous data, which affects the assessment of grid security and brings great hindrance to the automation of power system operation and maintenance. Therefore, we propose an AOA and fingerprint recognition-based sensing node location method to achieve accurate positioning of wireless sensing nodes to facilitate rapid troubleshooting. The method first enhances and denoises the signal characteristics of wireless sensing nodes to achieve pre-processing of wireless sensing fingerprint information and solve the problem of fingerprint feature recognition of data samples. After that we combine the node fingerprint information through deep learning models to achieve device type recognition of wireless sensing nodes. Finally, we design the AOA node localisation method, which uses the AOA coordinate system measurement model to perform distance measurement and coordinate conversion on the fingerprint features of wireless sensing nodes to achieve accurate and fast localisation of wireless sensing devices. The experiments prove that the method has a high accuracy rate in the identification and location of wireless sensing node devices, which can effectively improve the efficiency of grid automation operation and accurate troubleshooting.